Carbon in Guyana Guyana Guyana: Guyana : : : Preparing for REDD Preparing for REDD Preparing for REDD Preparing for REDD Implementation

(1)

Measurement and Measurement and Measurement and Measurement and Reporting of Forest Reporting of Forest Reporting of Forest Reporting of Forest Carbon in

Carbon in Carbon in

Carbon in Guyana Guyana Guyana: Guyana : : : Preparing for REDD Preparing for REDD Preparing for REDD Preparing for REDD Implementation

Implementation Implementation Implementation

by Jonas Cedergren country study

UN-REDD PROGRAMME

2009

(2)

The UN-REDD Programme, implemented by FAO, UNDP and UNEP, has two components: (i) assisting developing countries prepare and implement national REDD strategies and mechanisms; (ii) supporting the development of normative solutions and standardized approaches based on sound science for a REDD instrument linked with the UNFCCC. The programme helps empower countries to manage their REDD processes and will facilitate access to financial and technical assistance tailored to the specific needs of the countries.

The application of UNDP, UNEP and FAO rights-based and participatory approaches will also help ensure the rights of indigenous and forest-dwelling people are protected and the active involvement of local communities and relevant stakeholders and institutions in the design and implementation of REDD plans.

The programme is implemented through the UN Joint Programmes modalities, enabling rapid initiation of programme implementation and channeling of funds for REDD efforts, building on the in-country presence of UN agencies as a crucial support structure for countries. The UN-REDD Programme encourage coordinated and collaborative UN support to countries, thus maximizing efficiencies and effectiveness of the organizations’ collective input, consistent with the “One UN” approach advocated by UN members.

UN-REDD Programme contacts:

Peter Holmgren

Environment, Climate Change and Bioenergy Division

Food and Agriculture Organization of the United Nations (FAO) [email protected]

Tim Clairs

Bureau for Development Policy, Environment Group United Nations Development Programme (UNDP) [email protected]

Tim Kasten

Division of Environmental Policy Implementation United Nations Environment Programme (UNEP) [email protected]

Website: www.undp.org/mdtf/un-redd www.unredd.net

Disclaimer

The UN-REDD Programme MRV Working Paper Series is designed to reflect the activities and progress related to the Programme. These MRV Working Papers are not authoritative information sources – they do not reflect the official position of FAO, UNDP or UNEP and should not be used for official purposes.

The MRV Working Paper Series provides an important forum for the rapid release of information related to the UN-REDD Programme. Should readers find any errors in the documents or would like to provide comments for improving their quality, they are encouraged to get in touch with one of the above contacts.

(3)

1. Summary 1.1The Mission

The mission had three objectives. The first objective was to assess state and change of carbon pools including a review of data available and relevant work done. The second was to identify change processes – deforestation, conversion and degradation, including impact of logging practises on diversity and other ecosystem services. The third was to evaluate institutional capacities for MRV including institutional arrangements for MRV systems, governance and implementation.

1.2 State and Change of Carbon Pools

No biomass study has been conducted in Guyana.

Two assessments of carbon stocks and pools have been made using data from other countries (ter Steege 2001 & Schoch, D., et al. 2009). Differences between the studies are to a large extent explained by the different equations used.

Two studies are under way. One is an initiative from IUCN and the Guiana Shields Programme. No details of the study other than that it will be published early 2010 have been revealed (Chesney 2009. Personal Communication). The other is a large scale study biomass study carried out by the Guyana Forestry Commission (Alder and Kuijk 2009a).

Preliminary baseline calculations, mainly based on three more or less national inventories, have been made by the Forestry Commission study (Alder & Kuijk 2009b). Results from those calculation are those adopted in the present report.

Main results follow below:

Source: Alder & Kuijk 2009b

Inventory data from the early sixties and from 2003 suggest that an average loss of 640 square kilometres has been going on since the late fifties, corresponding to a loss of 0.4 per cent on a current area basis. This is a gross figure that needs to be investigated further.

Logging and shifting cultivation are at present assumed to cause no forest degradation.

Degradation is mainly caused by mining, expansion of settlements, infrastructure development and some, limited, conversion to agriculture. Road construction often cause hydrological changes and increased fire frequency.

Area Vegetation Class km²

Above ground biomass

Roots Necro- mass

Soil carbon Total

Above ground biomass

Roots Necro- mass

Soil carbon Total

Lowland Mixed Forest 100,408 180 40 19 33 271 1,810 398 190 326 2,725 9.99 Hill/Montane Forest 45,190 171 38 18 33 259 774 170 81 147 1,172 4.30 Wallaba Forest 10,867 230 51 24 33 337 250 55 26 35 366 1.34 High forest subtotal 156,465 181

40

19

33

272

2,834

623

297

509

4,263

15.63

Dakama Forest 4,234 92 20 10 22 143 39 9 4 9 61 0.22 Scrub or Savannah 17,562 38 8 4 22 72 67 15 7 38 127 0.46

Swamp/Marsh Forest 26,899

96

21

10

84

211

259

57

27

225

568

2.08

Cultivated/urban/cleared 4,687 - - - - - - - - -

Total land area 209,847

152

34

16

37

239

3,199

704

336

780

5,018

18.40

Carbon tonnes per hectare Total carbon, millions tonnes CO2

equiv.

Gigatonnes

(5)

The endpoint of deforestation/degradation is typically scrub/savannah with 72 tonnes of carbon per hectare. Total carbon loss is thus 200 tonnes per hectare, on a national level 12.8 million tonnes, assuming a “worst case scenario” of 640 square kilometres of deforestation per year.

1.3 Change Processes

Selective logging in Guyana is currently done at a level well within the realms of sustainability in terms of timber production and follows a stringent Code of Practice. Illegal logging occurs but at a relatively modest rate. It is difficult to see that selective logging would cause deforestation.

Repeated re-entries could cause forest degradation.

The effects of selective logging on biodiversity, soil and water has been given much attention in Guyana. Studies conclude that the physical environment as well as biodiversity can cope with logging as long as operational standards are high.

Shifting cultivation as done today in Guyana is on a small scale, and can be regarded as a balanced system. It is doubtful whether the shifting cultivation of today qualifies as forest degradation. It is highly unlikely that it causes deforestation. This could change to the worse if mining communities are formed or with other major development activities in the interior of the country.

A study in 2007 – 2008 found that 24 428 hectares had been cleared by mining, mainly in the northwest of the country where there is also a concentration of logging roads. Small and medium scale mining is most likely the chief driver of deforestation and degradation.

There is talk about Brazilian interests in establishing agricultural plantations in south Guyana. That is has yet to take place. There is land suitable for agriculture, so this is a scenario that may well materialise.

Forest plantations cover some 12 000 hectares (Anonymous 2005). There are no plans to expand forest plantations.

Sugar plantations have partly been converted to townships.

It cannot be ruled out that there are plantations that have been abandoned and “recaptured” by natural forest.

Infrastructure development makes areas available for development. Connecting Guyana with Brazil and making a large part of the interior available for various activities will of course have consequences for e.g. forestry, mining and agriculture.

1.4 Recommendations and Institutional Capacities for MRV

A REDD-reference level for Guyana must take assumptions on future development into account.

Natural resources must be declared to the degree possible. Forest land with high potential for other uses, e.g. oil palm plantations and hydropower, should be identified and mapped. An effort has already been done, but facts and figures need to be substantiated.

(6)

deforestation and degradation. Deforestation and degradation are further difficult to distinguish from each other. A calculated gross value has been used in the present study. This is an issue that need to be addressed.

Forest types relevant to carbon measurement need to be defined and studied. A baseline should contain a transparent statement of carbon stocks, and a reliable method to monitor their change over time. Change patterns in land use and their rates, and what implications ongoing land use processes have on greenhouse gas emissions. A mechanism to regularly report and verify ongoing development is also needed.

Forest management is practiced only on parts of the Guyana forest. Forest management however is not the only human activity applicable for forest land. Information on carbon stocks is thus needed from all parts of the country. Field data with national coverage is needed. The biomass study currently conducted by the Guyana Forestry Commission is an effort that will provide field data from all parts of the country. What is important is that a solid base of knowledge based on data collected on the ground is established for all parts of the country. Remote sensing is of limited value on areas barely inventoried in the past.

A problem for REDD is that there is at the moment no comprehensive land use plan for Guyana.

This means that the legal status of forest land is undefined. Several rights may apply to the same piece of land. This must be resolved if a REDD mechanism is to be launched.

Developments that typically follow improvements of infrastructure must not catch agencies concerned with surprise if a REDD mechanism is to be launched. Institutional preparedness is required. Inter agency co-operation will prove necessary.

The party paying revenue and the party receiving them must ensure that expectations are mutually understood.

An agency to work with REDD needs to be established. Such an agency should consist of staff from all agencies concerned by REDD. A council of representatives from the main groups of stakeholders to advice the “REDD Agency” would be helpful, as well as a steering committee made up of representatives for agencies involved, including the party paying the revenues. This would be helpful in handling and modifying expectations from parties concerned, and in making sure that all concerned are given opportunity to voice concerns, even after the launch of a REDD mechanism A successful launch of a REDD mechanism requires high capacity in a number of fields, e.g. land use planning, mapping, remote sensing, mining, forest biometrics, forest management. When planning for REDD, capacity building needs careful consideration.

Current adherence to forestry laws and Code of Practice needs to be evaluated. Work to ensure a high standard of forest management and to control illegal logging needs to continue. A capacity to address training needs as they are identified in the field needs to be put in place in combination with enforcement of regulations. Other forest management related factors that should be given special attention is forest management on smaller concessions, Amerindian land and other private land. The size and existence of a backlog of overlogged areas, i.e. areas logged before today’s practice was introduced, also needs to be investigated.

Small- and medium scale mining is an issue that must be addressed. In the meetings held it was often quoted as the foremost driver of deforestation and degradation. Environmental standards in mining should be documented and, if needed, improved.

(7)

A REDD mechanism will have effects beyond carbon. In general it will help install low carbon thinking in society. It will also be important in promoting co-operation between agencies and authorities. This will prove helpful also in non-carbon issues. A REDD mechanism will force the setting up of land use plans for areas concerned, and such plans are helpful also in non-carbon issues.

(8)

2. Summary of meetings

An important part of the mission in Guyana has been to meet with representatives of agencies involved or concerned by REDD. Valuable information, documents and comments have been obtained through these meetings. Below follows a summary of the meetings. A more complete account is found in Appendix 2 of the main report. Please note that opinions expressed are those of the individual(s) and agencies with whom meetings were held, not those of the consultant or FAO.

Guyana Forestry Commission.

Deforestation & Degradation. Deforestation is low. Mining is a main driver. Shifting cultivation is limited and balanced.

Changes. The commission will probably have to expand, e.g. to enable a National Forest Inventory.

REDD will probably require Guyana to improve resource utilisation and to invest in forest industry.

Comments. Forestry will continue on areas allocated as concessions even after the launch of REDD.

Forest Product Development and Marketing Council Incorporated

Problems. Worries about restrictions that may be brought about by REDD.

Comments. Given present logging rates, it is difficult to see that REDD should be a problem.

Forestry Training Centre

Changes. The institute may be utilised more as logging may have to become more professional Problems. Turnover rate among forest workers.

Comments. The institute could expand its activities without too much problems.

Chainsaw milling project (associated with Forestry Training Centre)

Changes. Even more important to improve quality of forest management by communities Problems. Communities are worried about restrictions they fear may come with REDD Comments. Demanding information campaign needed. Professionalism needed.

Department of Amerindian Affairs

Changes. Land use issues need to be addressed.

Comments. Communities need to diversify and develop their livelihoods and it is hoped that REDD will be helpful in this.

Iwokrama

Deforestation & Degradation. Mining is by far the most serious forest degrader in the country.

Changes. REDD will require more stringent stakeholder processes. It is essential that REDD is taken to the heart by rural population.

(9)

Problems. There is high pressure to get permits to log. Reduced impact logging is under introduction in the country as a whole, but will continue to face opposition from the loggers.

Comments. Vegetation mapping and ground truthing are expensive but necessary.

Guyana Geology and Mines Commission

Deforestation & Degradation. A mine is an example of disturbance, not degradation.

Changes. REDD mechanisms are welcome as long as they do not reduce people’s income. The commission is also strongly against deforestation. Mining should be done in a responsible way.

Problems. Little work done on site restoration after mining. Use of heavier machinery e.g.

excavators is on the increase, and they cause more damage. Some 250 excavators are active in mining.

Comments. Mercury is not seen as a problem.

Guiana Shields Project

Deforestation & Degradation. Mining. Extent and seriousness must be better verified. Hydropower projects discussed in the south. Savannahs could be converted to plantations. Brazilian companies interested in land for soy and rice cultivation. Oil could be found in South Guyana.

Changes. Lots of work is done by the Guyana Forestry Commission to raise operational standards.

Verification of progress in these efforts is greatly needed.

Problems. Guyana suffers from lack of baseline data. Another thing that is lacking is a legal status for forest land. A comprehensive land use plan is also lacking

Comments. Policy workers need to be trained. There is more or less no data at all from large parts of the country, the reason being that they have never been studied. A road from Georgetown to Lethem could have an impact on land use and carbon emissions.

Guyana Manufacturing and Services Association Limited

Changes. REDD will mean change in that rules will be implemented with greater stringency. Poorest practices have been checked by a stringent log tracking system, and regulations against land lording.

Problems. It is hope that a REDD mechanism will not reduce log production, which must already be considered low.

Comments. Things happening so fast that it is difficult to keep track, and formulate opinions.

Money distribution from REDD is going to be difficult. It is essential to get that right, from the very start.

Office of the President

Changes. The mechanism still needs to develop. No working model exists. It will be up to Guyana to pave the way.

(10)

World Wildlife Fund

Deforestation & Degradation. Mining is the main agent of deforestation. Mining practices are also detrimental to human health, mainly due to mercury pollution. Agriculture cannot at the moment be said to be a competitor to forestry. Shifting cultivation is to some extent.

Changes.Miners should be helped by prospecting to reduce unnecessary digging.

Problems.New technology for small scale mining has made practices even more destructive. Illegal and legal mining are equally destructive

Comments. Capacity building is a priority. Indeed REDD must have capacity as one of its components.

Conservation International

Deforestation & Degradation. The Georgetown – Lethem road, the driver of drivers. Unregulated mining. Intensive and unregulated logging. Ongoing efforts to shorten cutting cycles are a cause for concern. Shifting cultivation on the savannah is a danger to forest cover. pressure for land from Brazilian interests.

Changes. A major effort in awareness and information regarding REDD will be required for it to have any chance of success. REDD could instigate good governance.

Problems. REDD will force Guyana to review its natural resource management and land use planning.

Comments. Should REDD fail, there will be major implications for the forest of Guyana. How money is spent will also have implications. A top – down approach means that people may feel alienated.

REDD money must be enough to constitute an incentive.

Forest Producers Council

Deforestation & Degradation. Mining!

Changes.REDD has to be better understood before they are willing to take a firm stand on it. They do not know how much money REDD might give, and that is something that matters.

Problems. It is unfair that they should have to comply with high standards when miners can come and do as they please. They also feel that small scale loggers get away from the stringent standards prescribed by Code of Practice.

Comments. Guyana should make up its mind on the future of commercial forestry.

(11)

Amerindian organisations

Deforestation & Degradation. Great concern about mining among Amerindians. Shifting cultivation as presently done by Amerindians is rather benign in comparison to mining.

Changes. It seems that droughts and floods have increased. This has been particularly pronounced since 2004. New crops may have to be considered, and new land may have to be tilled.

Problems. The REDD process is little understood by the Amerindian community. There is the impression of a complex process that goes without the involvement of Amerindians.

Comments. Communication with Amerindians requires professionalism. Interpreters needed in consultations not only to have translation, but rather to have it translated into words that make sense to Amerindians. more information and time to comment on REDD. There is concern that REDD may mean restrictions on their land use.

University of Guyana

Deforestation & Degradation. Mining, shifting cultivation, expansion of communities and roads are what causes deforestation and degradation. Illegal logging was at its peak 1992 to 1994.

Changes. If REDD is launched, that will affect teaching in e.g. inventory and ecology. New courses will also have to be tailored to REDD. REDD will need a national forest inventory

Problems. REDD will require highly qualified staff.

Environmental Protection Agency

Problems. The role of EPA is unclear. Another thing that is unclear is how to distribute the money earned through REDD.

Comments. Less employment in forestry and mining will have implications for traditional life. REDD cannot be permitted to stop people from earning a living.

2.1 List of Tables

Table 1. Carbon content of the main soil- and forest types of Guyana Table 2. A summary of inventory results and biomass assessments.

Table 3. Pan-tropical biomass equation coefficients for major forest types Table 4. Soil Organic Matter to 1 meter depth

Table 5. Vegetation type areas and key to Map 2

(12)

Table 9. Ecosystem carbon including soil and necromass Table 10. Guyana forest areas in 1962

Table 11. Carbon stocks, losses, and retention in unlogged and logged tropical rain forests in Malaysia and Brazil. Carbon consequences of conventional logging and reduced- impact logging are compared

Table 12. Forest Allocation as Recorded by the Guyana Forestry Commission (December 2007) Table 13. A summary of plots established in plantations of Pinus caribaea.

2.2 List of Maps

Map 1. Inventory zones for FAO, CIDA and GFC surveys Map 2. Vegetation map of Guyana

Map 3. New roads and degraded forest area map of the State forest estate

Map 4. Logging Concessions, Forest Cover, and Amerindian Areas in Guyana

2.3 The mission in short

Together with experts and institutions in Guyana make preliminary assessments of forest carbon stock and changes, and evaluate institutional capacities for Measuring, Reporting and Verifying (MRV). Based on these findings make recommendations on the next steps to prepare a REDD mechanism. The complete Terms of Reference are found in Appendix 1.

2.4 Main issues to be covered

The mission can be broken down into three main parts:

1. State and change of carbon pools. A review of data available and relevant work done. Results will be analysed and evaluated

2. Change processes – deforestation, conversion and degradation. Work also includes impact of logging practises on diversity and other ecosystem services.

3. Institutional capacities for MARV. Options at different ambition levels, national and sub-national, should be defined. Institutional arrangements for MARV systems, governance and implementation investigated.

4.

Work has been based on a review of existing literature on Guyana’s forests and through contacts with relevant authorities and stakeholders and experts. Accounts of meetings held are found in Appendix 2, and literature consulted is briefly described in Appendix 3.

(13)

2.5 Context of mission

The United Nations Collaborative programme on Reducing Emissions from Deforestation and Forest Degradation in Developing Countries (UN-REDD Programme, www.un-redd.net) is a collaboration between FAO, UNDP and UNEP. A multi-donor trust fund was established in July 2008 that allows donors to pool resources and provides funding to activities towards this programme.

Based on request from Governments of Norway, Guyana and Surinam, FAO, as partner of UN-REDD, will field a consultant to work with institutions and experts in Guyana and Surinam to make preliminary national assessments of forest carbon stock and changes, evaluate the institutional capacities for Measuring, Reporting and Verifying (MRV) for REDD and make recommendations on next steps to prepare for a REDD mechanism.

(14)

3. State of knowledge and efforts under way on carbon pools in Guyana’s forest ecosystems

Biomass estimates are necessary for quantifying carbon stocks and pools. No direct estimates are available for Guyana. Nonetheless efforts to assess carbon stocks have been made, and there is work in progress. Below follows a review of work done, and work in progress.

The earliest is ter Steege (2001). Another effort is a project by Conservation International,

“Biodiversity Mainstreaming Through Avoided Deforestation, Guyana Case Study. Guyana Forestry Commission is conducting a study as parts of its effort to launch a REDD mechanism. The Guiana Shield Project is conducting a study in collaboration with IUCN.

Below follows an account of efforts done or under way. Sections are deliberately kept short to enable comparisons.

Preliminary results from the Guyana Forestry Commission study are used as estimates of carbon stocks and pools in the present study.

3.1 Materials and methods

Ter Steege (2001)

Ter Steege (2001) uses biomass estimates from Brazil, French Guyana and Venezuela. Data from Guyana emanate from an FAO project, covering eight of ten regions of the country. Data from central Guyana were used to develop stand tables per combination of soil type and forest type.

Data originated from the Great Falls inventory. Both inventories start recording tress at 30 cm dbh.

Lower size classes were constructed based on exponential decline, determined using regression.

Below ground biomass was estimated with a constant root-shoot ratio of 0.22 for mixed forest (Russell 1983, Jordan 1989). For forests on white sand root-shoot ratio may higher than 1. Large, woody litter was estimated to be 20 t/ha in undisturbed forest. Soil organic matter in the first 100 cm of the soil was calculated from soil carbon content of soil layers (van Kekem et al. 1996; with additional data from Khan et al. 1980, Gross-Braun 1965). The amount in the layer from 100-700 cm was calculated according to Nepstad et al. (1994) for forests in which deep rooting is expected.

Dead biomass (leaf and woody litter), may amount to another 150-200 t/ha in high forest. The standing stock of small litter (leaves, flowers, fruits, small twigs) is assumed to be in the order of 6 to 10 tonnes per hectare. Results are against this background often presented as a range rather than average.

Biomass carbon content was assumed to be 50 %, and soil organic matter had an assumed carbon content of 58 per cent.

(15)

Conservation International

The Conservation International project (Alexander, E.,et al. 2009, Killeen, T., et al., 2009 & Schoch, D., et al. 2009) based their biomass estimates on data from a CIDA supported national prism sampling project (the Interim Forestry Project) conducted 1990 to 1993, a management level inventory at Iwokrama conducted in 2003 and from permanent sample plots at Iwokrama established in 2007. The prism sampling was done using basal area factors of 2 to 9, with a minimum dbh of 10 cm for a tree to be included. The interim Forestry Project included 7943 sample plots along transects covering the country. Dbh of “in” trees was measured. The Iwokrama inventory consisted of 150 transects with 1451 plots in four forest types. Data from 13 permanent sample plots of 1 hectare were also used.

Dbh or diameter class midpoints were used to estimate live above ground biomass. The standard equation for moist broadleaf forest (Brown 1997 & Brown unpublished) was used with a correction factor for wood density. Wood densities are higher in Guyana than for the average tropical rain forest (Hammond 2005 & Baker, et al 2004). Special equations were used for species with exceptionally low wood density (Pourouma spp & Cecropia spp).

No direct measurements of root biomass are available for Guyana. The equation developed by Cairns et al. (1997) for upland forest was therefore used to estimate root biomass, it results in a ratio of 0.22 – 0.23 i.e. not unlike ter Steege. Field measurements of dead wood and litter were made using Alexander (Unpublished). After consulting studies made, it was concluded that forest floor litter carbon stocks do not surpass 2 % of live tree biomass carbon stocks. Soil carbon was assessed through a review of studies on the subject. Results vary considerably, maybe due to methodological differences.

Studies are in the process of being published. The consultant has been granted copies of drafts.

Thus, it cannot be ruled out that changes in estimates will be made.

Guyana Forestry Commission

A major effort to study biomass is under way at the Guyana Forestry Commission (Alder & Kuijk, 2009a). 900 permanent circular plots of 0.1 hectare, grouped as 180 clusters and 60 transects will be established nationally in all major forest types to measure and detect changes in forest biomass.

Trees over 20 cm dbh will be measured on the main plot, and over 5 cm dbh on the subplot (0.02 hectare). Plots will be organized in clusters of 5, based on a cross design with a central plot, and four plots on 100 m arms at right angles. Clusters will themselves be organised into transects of 3 clusters in a line, one km apart.

(16)

In addition data will be collected to establish coefficients and allometric functions for the major biomass pools. Two types of sampling units will be used. (1) Associated with monitoring plots will be 4 temporary 3 x 3 m quadrats which will be destructively sampled by weight for fallen deadwood, litter, and small plant biomass. Soil samples will be taken for organic carbon determination. These data will provide data that can be correlated with scores for deadwood and litter depth on the monitoring plots to derive biomass estimates for these carbon pools. (2) Detailed measurement of felled trees and tree roots to establish a data set of 300 plus sample trees for crown, bole and root biomass. Tree root excavations will be facilitated in cooperation with mining communities using hydraulic hoses.

Guiana Shield/IUCN

No details of the study have been revealed to the consultant other than that the work will be published early 2010.

3.2 Results and comments

Teer Steege (2001)

Results. A typical Guyanan forest was found to have an average carbon stock of 351 tonnes per hectare. This is made up of 150 tonnes from above ground tree biomass (including dead wood), 30 tonnes below ground, litter of 15 tonnes and 156 tonnes from soil organic matter.

Carbon stocks of biomass below and above ground including litter and dead lying wood were estimated for six soil types. Estimates ranged from 150 tonnes of carbon per hectare (laterite) to 195 (loam). Soil carbon storage estimates down to 800 cm soil depth ranged from 141 tonnes per hectare (white sand) to 490 (pegasse). Second highest was 163 tonnes per hectare (brown sand and loam). No measurements below 100 cm soil depth could be made for clay, laterite and pegasse.

Table 1 summarises carbon content over soil- and forest type.

Table 1. Total carbon content per hectare of the main forest types of Guyana Soil type Carbon content (tonnes per hectare) Mixed forest, brown sand 224

Mixed forest, loamy sands 260-358 Mixed forest, lateritic soil 286 Mixed forest, lateritic clay soils 321 Mora forest, alluvial clay soils 374 Swamp forest, pegasse soil 400-650 High forest, white sand 306

Low forest, white sand 67-306 dependent on site history

Source: ter Steege (2001)

Comments. Biomass is highly variable on a hectare basis. Thus, even if there were a few plots available in Guyana, they would only contribute to rough estimates. Reliable estimates based on total measurements will be expensive. Dbh is often the only information available in large scale data materials. Equations used to calculate biomass from Dbh have high explaining power on a tree by tree basis, but can give high errors when applied to a whole stand. There is considerable

(17)

disagreement between equations. That must be kept in mind. Regarding this study, it must be remembered that only trees with a Dbh bigger than 30 cm were actually measured, smaller trees trees were estimated based on regression assuming exponential decline.

Conservation International

Results. Data were grouped to permit estimates for the national forest classes defined by ter Steege (2001). Differences in live tree biomass carbon between forest types did not exceed 11 per cent. The lowest live biomass carbon found per hectare was forest on white sand in the prism inventory (188.9 tonnes per hectare), and the highest was 231.1 tonnes per hectare for swamp forest, the transect inventory of Iwokrama. For mixed forest, the comfortably most common forest type differences were below 5 per cent, ranging from 209.9 to 222.3 tonnes per hectare.

Lying dead wood accounted for 16 and 19 tonnes of carbon on brown and white sands,

respectively. No data was found that would help estimate carbon stocks in dead standing wood.

Litter carbon stocks were found to range from 1.6 to 3.4 tonnes per hectare. Data available on soil carbon range from 28 to 158 tonnes per hectare.

Comments. The prism data sets show greater variation, this is to expected. Many plots (or rather spots) may had four or fewer trees. ter Steege’s estimates are lower than these (17 to 34 per cent lower). The reason is that ter Steege used the Lescure et al (1983) equation which gives lower estimates through 95 cm dbh, and real differences in biomass stock in the materials. The Cairns et al. (1997)root biomass equation is a global average. It probably underestimates biomass on white sand, where large root systems are necessary to access water and nutrients. Assumptions regarding dead wood and litter are based on limited data.

Guyana Forestry Commission

Results are not yet available. Initial results from three plots will be made available in September 2009. A preliminary baseline on carbon stocks and pools has been calculated (Alder & Kuijk 2009b).

The plots, coefficients and allometric functions will give biomass estimates, together with

conventional forestry data on volume, species, forest and soil type, that are tightly geo-referenced and sampled over scales of 0.1 ha, 4 ha (clusters) and 2 km (transects). These will be used to supervise classification of LANDSAT imagery to determine areas and area changes of biomass and forest type cover classes. From this the monitoring system will be able to directly report carbon stocks and fluxes for REDD. This is the primary objective of the system. Its secondary objective is to provide a system of continuous national forest inventory providing strategic information on timber volumes and increments, NTFPs, biodiversity and other ecosystem services.

Guiana Shield/IUCN

Work will be published early 2010. Results are however reported not to differ dramatically from

(18)

4. Large-scale forest inventories in Guyana

Lacking domestic biomass studies, volume figures have to be relied upon for assessments of carbon stocks and pools.

Three major inventories have been made in Guyana. This section is mainly based on the work of Wright (1999) and Alder & Kuijk (2009b). When other sources are used, this is indicated in the text.

The first was done by FAO in the early seventies (Bratt 1971). Work covered the entire country, including the south. Clusters of four to five circular plots of 0.1 acres (0.04 hectares) were used.

Trees with a dbh of 1 foot (30.48 cm) or more were included. A total of 854 clusters were sampled.

The project resulted in a set of volume tables, grouped by species according to taper series (Bratt, 1971). Results are still used. This is the most comprehensive national inventory carried out in Guyana.

The second was a CIDA project from 1990 to1993. The project had a national coverage, with the southern parts of the country poorly represented. Fully stocked forests in the northern and central part of country were those mainly represented. There could thus be an overestimate of volumes.

Prisms with basal area factors of 4 to 8 were used. A Total of 7992 spots along transect lines were measured. The CIDA project further comprised measurements on 1849 felled trees from 137 species. There seem to have been no published models or tables as an output from the CIDA- project. The data was re-analysed by Alder (2000, 2001) and forms the basis of current volume calculation methods used by GFC for forest inventory and sustained yield calculations. Use of prisms with high basal area factors mean that typically only three to four trees would be sampled per plot (actually rather a spot). The large number of observations probably compensates for this.

The third is the management inventories done by the Guyana Forestry Commission since 2004.

Data has been collected on concession (thus not covering the southern parts of the country) using circular plots of 0.1 hectares. Poorly stocked forests are better represented than in the CIDA project. This is not a national inventory, nor is it meant to be. Work is concentrated on new production areas. A total of 2935 plots have been measured.

Prior to 1965 the colonial administration had undertaken numerous line surveys for individual concessions and permits, but there was no overall estimate of forest growing stock (Rees 1963).

Map 1 shows the coverage of the different inventory efforts.

(19)

#

##

#

###

## #

#

##

#

# #

#

# #

#

# #

##

####

#

##

#

# # #

#

Map 1. Inventory zones for FAO, CIDA and GFC surveys, purple dots depict FAO, red dots CIDA and green dots GFC.

Locations shown are estimated from place names and descriptions. GFC indicates centroids of inventoried areas.

Source: Alder & Kuijk (draft). Results of the inventories mentioned above are summarized in Table 2.

Table 2. A summary of calculated volumes per hectare and their distribution over diameter classes from the different large scale inventories in Guyana. Volumes are stated as gross above ground volume, over bark, 10 cm top, cubic metres per hectare by cm diameter classes.

Inv 0-9 10- 19

20- 29

30- 39

40- 49

50- 59

60- 69

70- 79

80- 89

90- 99

100- 109

110+ Total

FAO 57,06 46,26 42,38 34,07 17,34 9,27 5,90 5,38 12,99 230,64

CIDA 50,90 61,91 58,71 50,48 37,44 28,34 19,62 14,89 9,58 5,97 12,03 349,87 GFC 36,18 47,67 51,23 38,10 31,90 24,95 15,50 12,10 8,55 5,05 7,21 278,44 Source: Alder, Personal communication, 2009

(20)

5. Carbon Stocks and Their Change

Below follows an estimate of carbon stocks and pools in Guyana. It is highly likely that the estimate will be replaced by a more precise estimate once data from the ongoing GFC study becomes available.

The estimate is based on the work in Alder & Kuijk (2009). Other sources will be specifically indicated.

Conversion of Dbh to Volume

Gross bole volume is the over bark volume to the highest point of measurement, typically 10 cm top diameter. This is the appropriate volume to estimate expansion factors.

Using data of all species (1849 trees) from the CIDA inventory in a general regression on volume on diameter gives the following equation:

Ln V=-7.60027+2.2055ln(D), r2=0.92 {equation 1}

Transforming this equation to an unbiased volume equation requires Meyer’s (1944) correction.

This gives a final corrected equation of:

V=0.0005107*D2.2055 {equation 2}

This equation can be used to validate accuracy of pan-tropical equations for above ground biomass.

Estimating Above Ground Biomass from Volume

A number of approaches are available for Guyana. Below follows brief accounts of the options.

Brown (1997, §3.1.3) proposes an empirical equation in the absence of more detailed information.

This is an IPCC Tier 1 method that provides acceptable estimates when information is scarce.

ter Steege (2001) estimated average biomass stocking for Guyana using Lescure’s (1983) equation for French Guyana. Lescure requires no information on wood density or bole volume. Application in Guyana is therefore based on the assumption of similar species composition and stem form as in French Guyana.

Another approach is that of Chave et al. (2005). Data from 2410 trees from 27 study sites across the tropics. The study concludes that differences in the biomass equations between study sites are small if wood density variation is accounted for. The major significant factor is forest type.

Equations for dry, moist, wet and mangrove forests are produced in the study. These equations will be those used for the present report. The general equation is as follows:

W = ρ.exp(β0 + β1.ln(D) + β2.ln(D)² + β3.ln(D)³) {equation 3}

where W is tree above ground biomass in kg,  is wood specific gravity, D is tree diameter at dbh in cm, and the βi are coefficients which depend on forest type, as tabulated in Table 3 below. Note that the β2 and β3 coefficients are common across forest types, and β1 is the same for mangrove and wet forest.

(21)

Table 3: Pan-tropical biomass equation coefficients for major forest types

Tropical Forest type β0 β1 β2 β3

Dry. Over 5 months dry season, with marked water stress, rainfall below 1500 mm per year, may be semi-deciduous

-0.667 1.784 0.207 -0.0281

Moist. Marked dry season 1-4 months, rainfall 1500–

3500 mm per year

-1.499 2.148

Wet. Marked dry season 1-4 months, rainfall 1500–

3500 mm per year

-1.349 1.980

Mangrove. Coastal forests dominated by mangrove species

-1.239

Source: Chave et al, 2005

Dry and wet tropical forest types in Chave et al. models show lower biomass for a given diameter than the other equations. This is probably due to differences in average tree height on these less productive forest types. This emphasises the importance of correct forest classification especially relative to average height.

Below Ground Biomass

Below ground biomass is usually expressed as a root: shoot ratio relative to above ground biomass.

This is also implicit in IPCC (2006).

Brown (1997) found that the mean root: shoot ratio for tropical lowland moist forest is 0.12.

Higher values are found in ter Steege’s (2001) review of data from Suriname and the Brazilian Amazon. This is particularly true for drier forests and forests on poorer soils, where root systems need to be large to search for water and nutrients . For other forests a root: shoot ratio of 0.22 is assumed.

In the present report 0.22 will be used. Data for drier forests and those on poor soils is limited. This ratio may well be changed by results from the work outlined in Alder & Kuijk (2009).

Necromass

Necromass data is limited to studies from adjacent countries. Values range from 11.8 to 34.8 tonnes per hectare, with an average of 27. This is applied in this report as an expansion factor of 8.6 per cent.

(22)

Soil Organic Matter

Results are available for Guyana. Results are reviewed in ter Steege (2001). Results are presented in Table 4.

Table 4. Soil Organic Matter to 1 meter depth Soil type Tonnes per hectare Brown sand 65

Clay 167

Laterite 136

Loam 65

Pegasse 190

White sand 43 Source: ter Steege (2001)

Forest soils are assumed to hold 65 tonnes of organic matter per hectare, swampy and marshy areas 167 and scrub, savanna and Dakama formations 43 tonnes per hectare. These estimates are most likely conservative.

Biomass and Carbon Equivalent Ratios

The above approach would calculate biomass in tonnes per hectare. For carbon accounting this needs to be converted to tonnes of carbon per hectare. A conversion factor of 0.5 for biomass to carbon is used following advise in IPCC (Penman et al. 2003). For CO2-equivalent, tonnes of carbon are multiplied by the ratio 44/12, which is the ratio of their atomic weights.

Tier 1, 2 or 3

In terms of IPCC tiers (IPCC 2006), bole biomass are Tier 3, and the other components Tier 2.

Vegetation Classes and Forest Area

A vegetation map of Guyana (ter Steege, 2001) is shown in Map 2. Table 6 shows the related key to forest types and forest and other land use areas. These areas are calculated from the GIS shape file used in the map. The map itself is was developed by ter Steege from a variety of sources including satellite imagery, soil maps, research plots and forest inventory plots.

(23)

Map 2. Vegetation map of Guyana. Source: ter Steege (2001)

Vegetation map of Guyana

(24)

Source: ter Steege (2001)

Vegetation types in Table 5 were merged into broader vegetation classes. This is illustrated in Table 6. Table 6 also shows the distribution of plots after stratification.

Table 6. Creation of vegetation classes

Vegetation Class Vegetation Type (codes from Table 6)

Area Km²

FAO CIDA GFC

Cultivated /urban/cleared 9.0 4 687

Lowland Mixed Forest 1.1 to 1.4 & 2.4 100 408 548 5 169 2294 Hill/Montane Forest 1.5 to 1.7 & 7.1 to 8.1 45 190 241 145

Wallaba Forest 2.1 to 2.3 10 867 12 1 506 233

Dakama Forest 2.5 4 234 576 154

Scrub or Savanah 2.6 & 5.1 to 6.3 17 562 5 40 13

Swamp/Marsh Forest 3.1 to 4.1 26 899 48 407 56

Total 209 848 854 7698 2894

Waterways, open water 5 122

Guyana total area 214 970

Sample area (hectares) 171 123 289

Source: Alder & Kuijk (2009)

Table 5: Vegetation type areas and key to Map 2

Code Vegetation Types Map Area (km²)

1.1 Mixed forest Central/NE Guyana 20,858

1.2 Mixed forest NW Distict 28,393

1.3 Mixed Forest Pakaraimas 3,233

1.4 Mixed Forest South Guyana 47,789

1.5 Mixed Forest on steep hills 7,817

1.6 Mixed Forest on steep hills Pakaraimas 3,339

1.7 Mixed Forest on steep hills South Guyana 6,922

1.8 Mixed Forest/Swamp complex 2,513

2.1 Clump Wallaba Forest 1,016

2.2 Clump Wallaba/Wallaba Forest 2,522

2.3 Wallaba Forest 7,329

2.4 White Sand Forest South Guyana 136

2.5 Dakama Forest 4,234

2.6 Muri scrub/white sand savannah 3,810

3.1 Open Swamp 4,604

3.2 Marsh Forest 9,891

3.3 Coastal Swamp Forest 7,865

3.4 Forested Islands in Rivers 765

4.1 Mangrove Forest 1,262

5.1 Lowland grass/shrub savannah 11,287

6.1 Upland scleromorphic scrub 525

6.2 Upland grass/shrub savannah 1,940

6.3 Broadleaf upland meadow 196

7.1 Submontaine Forest Pakaraimas 23,549

7.2 Montaine Forest Pakaraimas 275

8.1 Submontaine Forest Southern Guyana 3,090

9.0 Clearings, cultivated land, large mines 4,687

Rivers, lakes, streams 5,123

TOTAL 214,970

(25)

Biomass and Carbon Stocks

Equations of Chave et al (2005) were used to calculate biomass on a tree by tree basis. Swamp and Marsh forests used the WET forest type equation, scrub and savannah used DRY type. Other types, i.e. most of the data used MOIST coefficients. Wood densities were found for the 134 most common species in Zanne et al (2009), for the remainder a weighted average of 0.71 was assumed.

Table 7 shows the biomass values calculated for all the inventories and vegetation classes by size class.

Table 7. Biomass estimates by size class for the FAO, CIDA and GFC inventories. Interpolated figures are shown in blue italic. Above ground biomass in tonnes per hectare by ten centimetre classes.

Source: Alder & Kuijk (draft)

The overall mean figure of 341 tonnes per hectare above ground biomass is weighted by plot numbers. Table 8 shows the areas for each vegetation class, above and below ground biomass estimates. The table also includes carbon an CO2 equivalents.

Inventory/Vegetation class0-9 10-19 20-29 30-39 40-49 50-59 60-69 70-79 80-89 90-99 100-109 110+ Total FAO 1968-73

Lowland Mixed Forest 20.0 ^31.4 ^48.7 57.4 50.3 51.5 46.3 25.0 10.4 7.4 5.8 6.7 360.9

Hill or Montane Forest 21.2 33.3 51.7 59.1 51.4 45.0 32.3 16.0 14.4 9.2 8.0 41.1 382.7

Wallaba Forest 19.5 30.5 47.4 78.5 67.9 70.9 26.1 10.2 350.9

Swamp/Marsh 11.2 17.5 27.2 25.1 25.2 27.5 28.4 11.9 11.1 3.8 9.1 3.6 201.5

Mean FAO 19.7 30.9 48.0 56.0 49.1 48.3 40.8 21.4 11.4 7.6 6.5 16.1 355.8

CIDA 1990-94

Lowland Mixed Forest 20.4 31.0 50.5 58.7 54.2 41.2 36.8 26.8 22.4 14.5 9.3 20.1 385.9

Wallaba Forest 24.0 35.0 60.9 74.4 81.6 73.4 47.1 33.4 21.8 12.6 7.5 11.8 483.3

Swamp/Marsh 10.6 15.8 26.7 28.2 28.2 21.5 17.6 10.8 11.1 10.7 6.9 11.5 199.7

Dakama Forest 28.5 60.8 53.0 28.1 14.3 8.4 4.3 2.2 0.7 0.4 0.3 0.3 201.2

Scrub or Savannah 10.8 28.1 15.0 12.2 5.3 4.3 1.3 1.5 78.6

Mean CIDA 21.0 33.4 50.5 55.7 52.9 42.2 33.9 24.3 19.2 12.3 7.8 15.8 369.1

GFC 2002-2008

Lowland Mixed Forest 15.9 23.4 40.2 48.5 40.3 36.5 30.3 20.9 16.8 12.5 7.0 11.2 303.6

Hill or Montane Forest 13.3 26.4 26.7 52.5 41.1 33.7 29.4 11.9 14.1 6.4 12.4 7.7 275.5

Wallaba Forest 14.7 24.5 34.3 59.3 55.1 47.2 36.6 17.9 10.0 5.7 4.8 1.1 311.3

Swamp/Marsh 11.5 19.9 26.1 29.6 17.6 12.0 7.9 2.0 1.2 1.3 129.1

Dakama Forest 17.2 32.4 36.3 22.1 5.3 3.4 0.8 0.2 0.5 118.2

Scrub or Savannah 7.1 13.7 14.7 20.2 8.6 1.5 3.0 68.8

Mean GFC 15.6 24.0 38.5 47.7 39.1 34.9 28.6 18.6 14.9 10.7 6.6 9.4 288.6

Mean, all inventories 18.1 28.7 43.8 53.7 49.3 40.8 33.1 22.7 17.6 11.6 7.4 14.3 341.1

Carbon in Guyana Guyana Guyana: Guyana : : : Preparing for REDD Preparing for REDD Preparing for REDD Preparing for REDD Implementation

Measurement and Measurement and Measurement and Measurement and Reporting of Forest Reporting of Forest Reporting of Forest Reporting of Forest Carbon in

Carbon in Carbon in